“Since statistics is applicable to almost anything, why not teach a class applying statistics to wine?” This was the question posed by student Wee Lee over an unmemorable bottle during a post-seminar reception back in 2005. Post-seminar receptions are common in academia. They are intended to encourage informal discussions inspired by the seminar, although I sometimes wish the wines were served before the seminar. The speaker might then seem or feel more inspiring, and at least one could blame the wine for any snoring heard during the talk. Wee Lee clearly was inspired by the intermingling of wine and statistics, even though the seminar was as unmemorable as the bottle.

How could I have said no to such an inspired idea, especially when I had just acquired half of a wine club’s inventory through a liquidation sale? As a pure academic, I was not then and am not now commonly given the resources or opportunity to acquire any fraction of any wine club’s inventory if the club owner has taken some statistics courses and understood the risk of running a business. However, one club’s loss is another academic’s fortune, the latter of which was increased by the wine club salesman’s appreciation of my curiosity about everything wine. “I’d rather give all the remaining bottles to someone who appreciates wine for two dollars a bottle than have them confiscated tomorrow,” he remarked. It was a sub-zero evening, but the salesman’s resentment of some prospective almighty enhanced the purchasing power of my wine budget by a couple of real zeros. It’s unclear if the “two dollars a bottle” qualification was his offer or valuation of my appreciation level. But either way, I joined – or rather closed – the club, in a serendipitous turn of events for a wine-connoisseur wannabe.

(No, you are not tipsy. Yes, this is a story of mixology of oenology and pedagogy. Pour yourself another glass and sip slowly to give me time to regale you.)

I, of course, could not possibly have kept all the fortune to myself. Lucky fortune cookies are meant to be shared, as are memorable bottles of wine – memories last far longer when there are multiple copies.

As a statistics professor, how could I have found a more appropriate use of my newfound fortune than to enhance the lovability of my beloved subject? In case you are not aged sufficiently to appreciate this rhetorical question, those were days when the answer, “I teach statistics”, was an effective turn-off line whenever I was too tired to converse with a taxi driver or a fellow passenger. That effectiveness naturally provoked me. How could statistics be taught in such a way that someday that line would be a turn-on?

Those were also the days when the concept of a wine cellar of my own was, well, just a concept. I had to move a car out of my garage during that winter when the salesman drove a car loaded with wine into it with almost flat tyres. I then moved the bottles into a basement cupboard as temperatures rose. In an attempt to control the temperature in the closet, I installed something that I would rather not mention for fear of embarrassing myself.

Wee Lee’s inspiration ultimately led to the creation of a brand new Gen Ed (General Education) course at Harvard – Real-Life Statistics: Your Chance for Happiness (or Misery). Traditional intro-level statistical courses arrange the content by mathematical and statistical difficulties and bring in stylised examples to illustrate how to apply formulas and carry out computation. The ‘Happy Course’ breaks with this longstanding, common practice, both literally and figuratively. The course offers six modules: Romance, Finance, Medical, Election, Legal, and Wine and Chocolate, made possible by my Happy Team, a group of graduate students who worked (and dined and wined) with me to develop and deploy the Happy Course.

During this course many pedagogical ideas were fermented over equally many bottles and reified through experiments conducted over several years. Statistical ideas and methods are brought in only when they are needed to address real problems. This led to the ordering and presentations of technical materials that would be unacceptable in the traditional framework.

And yes, there are plenty of students, and indeed faculty and deans at every university and college who consider maths to be synonymous with ‘aftermaths’. Teaching statistics as a mathematics subject has undoubtedly helped to turn an otherwise intoxicating subject into a conversation terminator, at least before the term Big Data became a headline.

As well being a fascinating topic to study from a statistical perspective, wine itself has beneficial qualities. Some of the most difficult conversations that I had to engage in with my colleagues, in my role as a department chair or graduate school dean, were made a bit more palatable by a bottle or two. Thus bringing wine into classrooms to ease some fear of maths is not a far-fetched idea, especially with the help of the Happy Team. However, despite knowing that when dealing with statistics you should always expect a bit of everything, I underestimated both the joy generated and the work involved in this singular adventure in statistical education.

Wine Tasting and Testing

Wine tasting provides a pedagogically engaging activity to demonstrate the essences and importance of experimental design, a gold standard for making causal conclusions, from clinical trials for treatment efficacy to safety assessments of autonomous vehicles. The myriad factors influencing wine quality and consumer preference require heedful designs to differentiate and distil them scientifically. Therefore, sharing my recently acquired wine collection via wine tasting was almost my first thought upon hearing Wee Lee’s proposal.

As it happened, the half collection I inherited consisted mostly of German Riesling, with every possible level of sweetness, from dry Kabinett to sweet Trockenbeerenauslese. Common wisdom has it that Riesling tends to be a favourite for newbies to wine, or at least the most easily accepted, because of its pleasing sweetness. Having introduced several colleagues and friends to the world of wine, my anecdotal observations supported that wisdom. However, anecdotes are not scientific data. If sweetness is an attractor, then it might induce different levels of preferences for the different types of Riesling. To test if this hypothesis is reasonable, one may conduct a wine tasting. But to make it scientific, one must adhere to several principles of experimental design. The obvious one is that it needs to be a blind tasting, just as clinical trials need to be double blind (i.e. neither patients nor doctors are informed about which patients receive which treatments) whenever feasible. This principle of blindness is to ensure that our scientific vision is not blurred by our judgement vulnerability, such as to label information or price tag. Wine tasting perhaps has the most dramatic and consequential testimony to offer in support of this principle. You would have to consume many, many bottles to imagine that the 1976 Judgement of Paris would have reached the same verdict, which revolutionised the wine world, if it had been conducted unblind.

The principle of randomisation, however, might be less obvious because many people equate randomness with haphazardness. In experimental designs – and in all statistical theory and methods – randomisation is the opposite of a haphazard process because randomisation means the process is under human control and we know precisely what can happen and how frequently it happens. A typical randomisation requires that everyone, or every possibility, be given an equal chance. For comparing different wines, the order of tasting can influence our judgement. (“Never serve your trophy as the third bottle” was advice given to me by a fellow fan of both wine and statistics many years ago over a most memorable lunch.) A well-designed tasting experiment, therefore, will assign approximately the same number of tasters to every possible order. In the Happy Course Riesling tasting, we served three kinds of wine (with all bottles wrapped), resulting in six possible orderings. We had 23 tasters, and hence each order was given to four tasters except for one – see the accompanying artistic rendition of a presentation slide.

(It is probably a good test to see if you need another glass, depending on whether you can immediately tell which order was the exception.)

The slide was from the actual lecture, where a table documented the average rating for any wine-order combination. I will refrain from sharing the rest of many slides for fear of losing your interest. But if your glass still is half full, I’d invite you to pair the half glass with a half number game: what can I conclude from these numbers? Let’s see. Looks like there is an ordering effect, since the two highest average ratings all occurred in the first row. But then the second lowest average rating also occurred in the first row. Wait. The whole Spätlese column received the three lowest ratings, and the first and second average ratings are much closer than that of Kabinett or Auslese. Perhaps then it is okay to declare an ordering effect since these averages were all based on a handful of tasters, and hence we should permit some degrees of give-or-take? But what degrees are acceptable, and how would that be determined?

Well, that’s why I invited you to play only a half number game, because these averages don’t tell the whole story. How much give-or-take should be allowed will depend on how individual ratings differ from the averages (or equivalently from each other). The more they differ, the more give-or-take the results are; larger individual differences as observed from the tasters suggest rather different preference ratings if we had different tasters, or even the same 23 tasters with the same tasting but for a different ordering assignment. Hence, we need to give ourselves larger slack to reduce our overconfidence from a single experiment, however scientific it might be.

In case my stat preaching is getting you dizzy (instead of tipsy), let me stop here to say that the actual statistical analysis is simpler and less confusing than ad hoc number games because they follow well-specified probability rules and can be carried out by computer. As a matter of fact, the whole idea of statistical analysis (and more broadly data science) is to help humans navigate mind-boggling ‘number games’ created by complex problems – judging wine quality and understanding consumer preferences is one of them. We can then see the big picture and act on fundamentals, instead of getting lost in a maze or jungle. Through pedagogical activities such as wine tasting, the students on the Happy Course got a direct taste of how real-life statistics helps reveal simplicity within and from complexity, a process, perhaps ironically, not unlike seeking complexity in seemingly simple fermented grape juice: both processes require training, practice, understanding, judgement, and a bit of luck.

Chocolate (wine) is happiness that you can eat (drink).

Astute readers may be wondering about a complication I’ve deliberately not yet mentioned: how could we serve wine to students under 21? The short answer is that we couldn’t and wouldn’t. The longer story requires another glass, and a box of chocolates, if you want to double your indulgence.

Wee Lee did remind me of the possibility of obtaining an education waiver of the age limit. For a popular course on wine tasting that he took from a school of hotel management, the course was permitted to conduct wine tasting for under-aged students as long as they promised to sip but not to swallow. Whereas I certainly should trust students’ sip-but-not-swallow promise no less than a US president’s smoked-but-didn’t-inhale confession, I had no confidence that such a waiver would be granted to a statistical course, no matter how charming or persuasive its instructor might be. Besides, not everyone enjoys wine as aqua vitae, nor should alcohol be foisted upon anyone at risk of allergic or otherwise adverse reactions, physically or spiritually.

Fortunately, my creative Happy Team came up with a tasty solution: chocolate tasting for those who cannot or won’t touch alcohol. It is particularly fitting for the Happy Course, since, after all, “Chocolate is happiness that you can eat” (Ursula Kohaupt).

The creativity of the Happy Team did not stop there. Once thinking about chocolate, they wondered if they could double the happiness by devouring both wine and chocolate concomitantly? This happy thought led to an experiment in chocolatier-ing and pedagogy, which in turn created the most unexpected and cogent teaching moment in my entire journey as a teacher.

The chocolatier-ing experiment was carried out by my teaching fellows, who learned from scratch how to make chocolates, with and without a wine-filled centre. The pedagogical experiment randomly mixed the two types of chocolates (one third wine-laced and two thirds without), and then invited 20 students (who didn’t mind being doubly happy) each to devour one chocolate and report if it was filled with wine or not. The resulting percentages were displayed after everyone responded. The experiment was repeated for another 20 students, but with the percentages from all 40 students computed and displayed; and then again for another 20 students, with a total of 60 responses accumulated.

The experiment was intended to illustrate three key notions in statistical estimation. The first is survey sampling; that is, learning reliably about a large population from a small representative sample. The second is the law of large numbers, which may be as intimidating to some as the law of the state. But all it says in a practical term is that the more chocolates consumed, the closer would be the reported percentage of chocolates injected with wine to one third (which was unknown to the students). The third is the confidence interval, commonly constructed as the estimate plus or minus the margin of error. This corresponds to having 95% confidence; that is, in repeated experiments, the interval so constructed should cover the truth 95% of the time. Here the margin of error matches, conceptually, with the common notion of the amount to give or take. In this case it is just slightly less than the reciprocal of the square root of the sample size, and hence it varies from about 0.22 to 0.13 as the sample size changes from 20 to 60.

Real-life statistics helps reveal simplicity within and from complexity, a process, perhaps ironically, not unlike seeking complexity in seemingly simple fermented grape juice: both processes require training, practice, understanding, judgement, and a bit of luck.

These theoretical notions assume that our experiments or data behave just as we design. The double-happiness experiment reminded us once more that theorised happiness could be real-life misery if we naively mix wishes with reality. When none of the first 20 students reported having consumed a wine-centred chocolate, my panic kicked in. Even without understanding the confidence interval calculation, what would be the chance that none of the 20 students got a wine-filled chocolate when supposedly each would have 33.3% chance of enjoying such a dose of double happiness? The theoretical answer is ²/₃ to the 20th power, which is about 3 per 10,000.

Events with very small probabilities (such as winning a lottery or being struck by lightning) of course can happen on a lucky or unlucky day. But when they happen, a healthy dose of suspicion that something went wrong is always wise.

The first thing to check was if, somehow, we hadn’t put in one third of wine-filled chocolates or we hadn’t mixed them thoroughly with the unfilled ones. Upon double checking, both possibilities were ruled out, and we further mixed the chocolates before the second replication. But this did not correct the problem, and indeed by the time we had all 60 reports, only 17% of students reported having consumed wine-filled chocolates, which would lead to the 95% confidence interval being from 4% (=17%-13%) to 30% (=17%+13%), which clearly missed the 33.3% rate by design.

So, what went wrong? The answer became clearer when some students complained that it was hard to tell if a chocolate was filled with wine or not. I immediately realised that I had been given a perfect case for teaching the concept of ‘control’ in experimental design. A critical design flaw with the double-happiness experiment was that each student was only given one chocolate. Without a regular chocolate to serve as a control for the purpose of comparison, many students who got the wine-filled chocolates were unable to detect the extra wine flavour, given that the minuscule amount of wine was surrounded by an overpowering amount of chocolate. This is essentially the same reason why a clinical trial requires a control treatment, often a placebo, for the comparative purpose of determining if there is an extra efficacy attributable to the experimental treatment only.

An embarrassed teacher might have claimed that the flaw was by design for pedagogical purposes – and indeed, the students might not have been able to tell because my shifting from discussing the sample survey to experimental design seemed seamless. But the truth is that I had failed to practise what I preach. The experiment was originally designed to illustrate a sampling survey, not experimental design, which was the subject for more wine and chocolate tasting events scheduled at the end of the semester. Nevertheless, good design principles apply everywhere. A simple pre-comparative tasting would have revealed the problem, but somehow it just never occurred to me. Ironically, retrospectively I felt fortunate to have been presented with the embarrassing design flaw, because if we had realised it earlier, we might have nixed the double-happiness experiment altogether and lost the great pedagogical lesson. This is because it was impossible to find any wine sufficiently pungent to compete with the rich chocolate without a control comparison. But implementing a controlled experiment would have been too time-consuming (and potentially confusing) in the amount of class time we could allocate to it.

However, ultimately we did find a solution, after desperation compelled a change of spirit(s): whisky, to be precise. Teaching the Happy Course opened up a club of whiskies for me, and this time they were all zero buck chucks.

Whisky: uisGe beatha?

For the purpose of reifying the modules even more for students, I was keen to enlist a speaker who had practised what they preached or preached what they practised. Since the module on wine and chocolate was the last one, this speaker would also serve as the grand finale speaker of the Happy Course. As such, we wanted to find a most memorable one. At that time, being a newbie to wine, I knew too little about the wine community to even come up with a single suggestion. However, years before, I had come across an article on classifying whiskies by statistician and whisky connoisseur David Wishart. It was just a curiosity then, but the article stayed in my memory, partially because any qualified statistician would wonder if the author was somehow related to the John Wishart of the Wishart distribution, a household term in statistics. In fact he wasn’t related, but David did accept our invitation.

As we excitedly anticipated his visit, something unexpected happened. Various different-shaped bottles of whisky with all kinds of labels arrived at my office, daily, to a point that my department got worried and inquired about their sources. I had no idea other than that it must have something to do with David’s visit. Indeed, as we learned later, David called many distilleries, informed them of his upcoming visit, and asked them to send me some samples. It was a clear testimony to David’s influence in the world of whisky. I ended up receiving 40 bottles. At the time I had little idea about which was which. But I sensed the attractions that they held for whisky aficionados, especially when paired with David’s presentation. We therefore advertised the grand finale lecture as a general seminar, with a whisky tasting to follow.

David’s lecture in December, 2010 gave an overview of a rich history of whisky, from medieval monks distilling beer to make whisky through to Prohibition-era doctors prescribing whisky for medicinal purposes. It also introduced novices like me to the world of whisky making – I no longer had to pity myself for not understanding the peating process. Most importantly, David showed how he had analysed nearly 1,000 tasting notes containing over 500 terms describing single malt whiskies. The analysis resulted in 12 cardinal flavours (body, spicy, sweetness, winey, smoky, nutty, medicinal, malty, tobacco, fruity, honey, and floral); a great demonstration of gleaning simplicity from complexity. This 12-flavour categorisation, together with a 5-point scale (not present, low hints, medium notes, definite notes, pronounced), enabled the clustering of Scottish whiskies into groups of similar ones, which “helps you choose a single malt whisky that suits your palate, not someone else’s”. (Much of David’s lecture was nicely summarised in his article on The Flavour of Whisky which appeared in Significance, a statistical magazine for the general public.)

The 40 bottles provided the most engaging post-seminar reception for many attendees, who got to taste and test David’s classifications. Of course, 40 bottles of fine whisky couldn’t possibly be consumed over one post-seminar reception, even with many colleagues attending from other departments. Thanks to the longevity of whisky, this engaging event was repeated multiple times in the subsequent two years until I was appointed dean of the graduate school of arts and sciences. These events clearly made an impression on their attendees, one of whom wrote to me saying: “I know how you became a dean – you bribed everyone!”.

By then my teaching fellow Casey, who had helped me run the Happy Course, had graduated and had become an instructor herself, so naturally she took over the Happy Course when my schedule no longer permitted me the time. When she contacted me to suggest a grand finale speaker, I told her to try David again. David responded to Casey’s invitation again most promptly, and the trip was confirmed for November, 2012. Lo and behold, just like two years earlier, Casey started to receive bottles daily and this time there were 41 bottles in total.

David’s 100-slide presentation that November was an eye-opener to many who didn’t expect that the nerdy subject of statistics could be this intoxicating. The tasting event that followed (that neatly doubled as the Dean’s Seasonal Toast), was a mouth-opener to some who had come primarily for the lecture but then stayed to taste some of the whiskies. A sip or two of The Macallan 25 Year Old was all it took to change their conviction (or claim), though frankly I got a bit worried that soon they would also get a taste of regression towards the mean.

Completely unexpectedly, the post-lecture tasting event and dean’s toast turned out to be far more romantic than planned. Just as The Macallan 25 Year Old was disappearing rapidly so did the light – thanks to a power outage. Several staff members rushed to Harvard Square to buy candles and lanterns and the tasting continued in soft lighting and to the accompaniment of Chopin. The tasting was in a graduate common room with a grand piano, and a talented student attending from my department was also a concert pianist. When the mellifluous Chopin meets a mature Macallan, surrounded by candles and lanterns, a romantic mood is almost mandatory.

Word got out quickly that the dean was a whisky aficionado and that his seasonal toast should not be missed. Appointments at the dean’s office also became popular – not least because of the embarrassment of whisky riches residing under my desk (and fortunately, no thermodynamic knowledge to keep them at optimal condition was required this time).

This is exactly how I imagined an educational holiday party should be.

My epicurean and pedagogical adventures seemed to be all coincidental and serendipitous, but ultimately, they led to an organised effort during my deanship thanks to a leader in the world of wine, and a supportive staff team. Every (new) dean needs some initiatives to make their presence known. Mine were: professional development for students, global engagement for alumni, and fundraising for the school.

Naturally global engagement and fundraising called for many trips, which offered me ample occasions to test the common wisdom that wine is a social lubricant. The trips also gave me the opportunity to meet many extraordinarily rich individuals, whether entrepreneurially, intellectually, spiritually or financially. One of them was Don St. Pierre, then the co-founder of ASC Fine Wines. Our conversation quickly turned into our shared passion – teaching. Don was eager to help young talents with his singular experiences from adventuring on foreign soils, literally and figuratively, including venturing into the wine business without any background in wine. His accomplishments earned him many accolades, epitomised by being elected as the International Man of the Year by Wine Enthusiast magazine in 2011.

An impactful professional development programme should help students experience the world outside of academia before they live in it daily, by choice or by necessity. Connecting students with industrial and business leaders is an obvious step. Don’s passion for sharing his experience then is a perfect match. I was therefore delighted when Don became an Entrepreneur-in-Residence for Harvard’s Innovative Labs (i-labs), an ecosystem that supports ‘Harvard students and selected alumni in their quest to explore the world of game-changing innovation and entrepreneurship’. Don’s presence on Harvard’s campus also elevated the dean’s seasonal toast from should-not-be-missed to must-not-be-missed.

During his trip to i-labs in Dec 2013, he brought some iconic Australian wine (Penfolds Grange and Bin 51 Rieslings, and Leeuwin Estate Art Series) as well as Master of Wine Sandy Block (of Legal Seafood). Together with my able staff team, the pair turned the graduate school common room into a professional tasting room. Don started the event with an educational presentation: ‘Behind the Grapes: Learning Australia’s Distinct Wine Regions’, interlaced with some of his own adventure stories. With ample food for thought, the audiences were then expertly guided by Sandy to prepare the wine for their palates, or rather to train their palates to best appreciate some very fine wine undoubtedly beyond their spending comfort zone.

With everyone merry (me especially), my seasonal toast was as expressive and lingering as the 2007 Grange. I don’t recall a single sentence but can still hear the laughter and applause, and a faculty member’s endorsement: “This is exactly how I imagined an educational holiday party should be.”

Greatly encouraged, Don and I started to plan for future seasons almost immediately after. Since we had started with Australian wine, it seemed quite natural to continue the alphabetic theme, with B for Bordeaux or Burgundy.

For the second event, Don brought in Prince Robert of Luxembourg, president of the Domaine Clarence Dillon estate, which oversees production of Château Haut-Brion, Château Tertre Daugay, and Château Quintus, among other brands. With a fully packed room in December 2014, Prince Robert delineated his family’s history with Haut-Brion, and then led a tasting of a dozen wines representing the super-premium offerings of Domaine Clarence Dillon and Bordeaux – from a 1995 Château Haut-Brion to a 2012 La Clarte De Haut-Brion. It goes without saying (or drinking) that many attendees felt that they were daydreaming, sipping Haut-Brion with – and offered by – an actual prince.

Daydreaming or not, I worried how we could follow the event. But Don did it again, and this time he brought not one speaker but two: Bill and Will Harlan, the father and son team from Harlan Estate – C is for California (sorry, Champagne!). And the pair’s presentation inspired even larger dreams. As Bill Harlan, the father, wrote in his poetic ‘A Note from the Proprietor’, “Every life is full of dreams … Those of us who awaken at some point to the desire for something more lasting than momentary pleasures are lucky in the best sense of the word. That knowledge – the awareness that growth cannot occur in the absence of roots – makes it possible to imagine, and perhaps to pull into existence, something that may last for many generations.” I was particularly intrigued by the Harlans’ discussion of thinking and business planning on the scale of 150 years. Why 150 years? Bill’s answer still stays with me: 150 years represent the typical generation span which still permits a direct memorial lane or knowledge path, from one’s grandparents to one’s grandchildren.

Whether driven by the vine or the divine, the tasting event is cyclingback to its first growth, blending wine and statistics, while the statistical discipline itself is being fermented into a grande cuvée of data science.

The dream continued in December 2016 with the theme D for Diversity, showcasing a Chinese red, a New Zealand white, a you-know-where Champagne and even a Cognac, all offered by Moët Hennessy’s Estates & Wines. A Chinese red from Moët Hennessy? Yes, Ao Yun, a blend of 90% Cabernet Sauvignon and 10% Cabernet Franc grapes grown in a perfect mix of sunshine and shadows cast from the foothills of the Himalayas, near Shangri-La (formerly known as Zhongdian County Town in the Yunnan region). Making China’s first luxury wine that can compete with the worlds’ greatest wines is a dream in progress, with its challenges as epic as its mission. “How could one build a world-class winery without electricity?” asked Jean-Guillaume Prats, then the CEO of Moët Hennessy’s Estates & Wines, whose presentation that December was as captivating as the Ao Yun wine itself.

But where would E take us in the wine world? By then (December 2017) I had completed a five-year deanship, which earned me a much-needed sabbatical – meaning I would have uninterrupted days to seek simplicity in statistics, complexity in wine, and everything else in fortune cookies. As luck would have it, my successor Emma Dench is a professor of classics, and that Don arranged a tasting of Kir-Yianni and Sigalas wines from Ελληνική Δημοκρατία (Hellenic Republic, aka Greece). This perfect pairing was not only reverberated by a double E, but more importantly provided the audience with another double treat: a singular tour, led by a world-class scholar in classics, of the historical and cultural context of wine in ancient Greece, and a tasting of the award-winning wines led by a representative from the estates of Kir-Yianni in the north-east of Greece, and Sigalas from the island of Santorini.

The following year, Don brought in yet another award-winning wine, together with its producer; in December 2018 Emma’s Dean’s Seasonal Toast featured Gaja wine, a tasting led by someone who many regard as ‘the undisputed king of Barbaresco’, Angelo Gaja. The elegant and opulent wines were paired perfectly with an energetic, wide-ranging presentation by the charismatic Gaja, who covered topics from the impact of climate change on wine to his philosophy and practice of teaching and transmitting knowledge.

But hang on – what happened to F? Fame? Finest? Fantastic? Unfortunately, the silent F turned out to be Finale. By 2019, I had immersed myself in Harvard Data Science Review (HDSR), a flagship publication of Harvard Data Science Initiative (HDSI). Emma and I talked about moving the wine tasting to HDSI, and Don and I started to discuss the possibility of holding an annual workshop on data science for wine, which would be a natural occasion to continue the tasting event. Then the world changed its natural course thanks to COVID-19. Many activities found virtual substitutions that were at least acceptable. But wine tasting is not one of them. Wine cheering over Zoom can never match wine sharing in a room.

Coincidence is god’s way of remaining anonymous.

Was it merely a coincidence that the Dean’s Seasonal Toast ended on F? According to Albert Einstein, “Coincidence is God’s way of remaining anonymous.” He would have been a great fortune cookie writer. (Yes, there is such a profession.) Or, as Bill Harlan put it, “Everything happens in season … It is only in retrospect, however, that we can see how our lives come to be what they are, no matter how well or wisely or far ahead we think we plan them.” Reflecting on the journey from A to F (over a glass of 1999 Chateau Musar), my lingering turned into longing. The seasonal toast was meant to be only for a season. It was on its way for its next season, pairing wine with data science, an artificial ecosystem, regardless of how the natural ecosystem evolves. Whether driven by the vine or the divine, the tasting event is cycling back to its first growth, blending wine and statistics, while the statistical discipline itself is being fermented into a grande cuvée of data science.

As a data-driven fortune teller, I’m eagerly seeking data and signs, from fortune cookies to fortune/future makers. Do I have any doubt that Harvard will once again find a way to pair wine and data science? Of course, I do – no fortune teller can tell their own fortune. But every time I leave my now properly temperature-controlled ‘meditation’ room, the doubt lessens, thanks to a souvenir from a post-seminar trip to Napa Valley, organised by my statistical colleagues at University of California, Davis. Apparently, as members of a university known for its viticulture and oenology studies and degree programs, my colleagues there really understand the seductive and inspirational power of wine. It was an engraved barrel stave, welcoming anyone entering my meditation room: When in doubt add more wine.

This article was written by Xiao-Li Meng, the Whipple V. N. Jones Professor of Statistics at Harvard University, and the Founding Editor-in-Chief of Harvard Data Science Review.

Originally published in FONDATA, Issue Three.